bibkey: ganardi2022hierarchy authors: Ray Ganardi; Marek Miller; Tomasz Paterek; Marek Żukowski year: 2022 title: “Hierarchy of correlation quantifiers comparable to negativity” doi: 10.22331/q-2022-02-16-654 url: https://arxiv.org/abs/2111.11887v2 claim: “For a bipartite state rho with partial transpose rho^{T_B} on the second system, the negativity is N(rho) = (||rho^{T_B}||_1 - 1)/2 and the partial transpose distance is d_T(rho, sigma) = ||rho^{T_B} - sigma^{T_B}||_1 / 2. Conjecture 1: for every density matrix rho, the infimum of d_T(rho, sigma) over the PPT states sigma equals N(rho).” strata_touched:
- D5/S3/Quantum/Entanglement/NegativityPartialTransposeDistanceRefutation license: citation-only triage: anchor
Hierarchy of correlation quantifiers comparable to negativity
Ray Ganardi, Marek Miller, Tomasz Paterek, Marek Żukowski, Quantum 6, 654
(2022); arXiv:2111.11887v2 (quant-ph). Quotations are from the arXiv v2
source, with its macros kept as printed (\pt{\rho} is ,
\norm{A} is , \dm is the set of density matrices on
).
The partial transpose and the trace norm:
Given a bipartite system
\hs_A \otimes \hs_B, we denote partial transpose on the computational basis of\hs_Bas .
The trace norm of a matrix is defined as
\norm{A} = \Tr{\sqrt{A^\dagger A}}.
The PPT states and the negativity:
A state is PPT if
\pt{\rho} \geq 0, and we denote the collection of such states as\PPT.
The negativity of a state is defined as
N(\rho) = \half \left(\norm{\pt{\rho}} - 1\right).
The partial transpose distance (Definition):
Let
\rho, \sigma \in \dmbe two density matrices. The partial transpose distance is defined as \begin{align} d_T(\rho, \sigma) &= \half \norm{\pt{\rho} - \pt{\sigma}}. \end{align}
Conjecture 1:
Let be a density matrix. Then \begin{align*} \inf_{\sigma \in \PPT} d_T(\rho, \sigma) &= N(\rho). \end{align*}
The paper proves the equality for states of positive binegativity, , and reports that random mixed states of two qudits for each show no counterexample.
The counterexample state of the settlement belongs to the punch-card family of L. Lami, F. A. Mele, B. Regula, arXiv:2405.09613v2 (Phys. Rev. Lett. 134, 090202 (2025)), Eq. (15), where the family is used for the additivity of the κ-entanglement; the separation of the PPT distance from the negativity is not stated there.
The encoding takes states on , indexes
matrices by Fin d × Fin d, transposes the second factor, and uses the
trace norm ; the
negativity and the distance are written in the claim exactly as defined
above.
Verified locator
- DOI: https://doi.org/10.22331/q-2022-02-16-654 (Quantum 6, 654 (2022)).
- URL: https://arxiv.org/abs/2111.11887v2 (v2, 2022-02-09, the latest
version; source
main.tex, md5df386f803c4233dab0a6307e9c4dc578): the partial transpose (l. 153), the trace norm (l. 155), the PPT states (l. 159), the negativity (l. 163), the partial transpose distance (l. 172–177), Conjecture 1 (l. 268–274) and the numerical evidence (l. 316–320).